Maintenance decisions for high voltage assets are usually based on experience and failure statistics, however often too low sample sizes are available. On the one hand failure data do show that aging processes proceed inevitably in network components during a certain period of time before failure. Therefore diagnostic tools to assess the condition of high voltage components are very useful when the overall age of the electrical infrastructure increases. Depending on the diagnostic tool used, the results of the measurements involve different diagnostic parameters, e.g. partial discharge (PD) inception voltage, PD magnitudes at different voltages, tangent delta and return voltages. It is also known that aging and degradation, as failure mechanisms, are statistical in nature. On the other hand the amount of diagnostic data can be large, enabling an additional source of statistical information to estimate the time to failure more at specific components level. Due to stochastic behavior of the underlying degradation mechanisms the diagnostic data can be statistically interpreted. As a result boundaries for condition indexes for good, moderate and bad technical performance are derived, which are more adequately for implementation in an diagnostic decision support system of a utility, in particular to direct maintenance and replacement policy of components.